Mapping the causal structure of AI governance factors influencing quality assurance under Industry 4.0 regulations

Purpose The increasing deployment of artificial intelligence (AI) in Industry 4.0 has revealed limitations in traditional quality assurance systems, which are not designed to govern algorithmic decision-making and autonomous operations. This study aims to empirically examine how AI governance dimensions causally influence quality assurance outcomes in AI-enabled industrial systems. Design/methodology/approach A Delphi-based expert elicitation involving 10 specialists from industry, technology, policy and regulation was conducted. The fuzzy decision-making trial and evaluation laboratory (fuzzy DEMATEL) method was applied using linguistic variables and triangular fuzzy numbers to model cause-and-effect relationships among governance dimensions. Findings The analysis identifies a clear causal hierarchy. Algorithmic transparency and cybersecurity assurance emerge as dominant causal drivers, supported by data governance quality, continuous improvement culture and integration of ISO and AI standards. Ethical accountability, regulatory compliance readiness, stakeholder trust, process reliability and top management commitment appear as effect dimensions. Strong feedback loops indicate non-linear governance maturity and cascading effects. Research limitations/implications This study empirically maps the causal relationships between AI governance and quality assurance using fuzzy DEMATEL; however, several limitations should be acknowledged. First, the expert panel comprised 10 professionals primarily from highly regulated sectors in developed economies, which may limit generalizability to SMEs, emerging economies or less-regulated industries. Second, the fuzzy DEMATEL approach relies on expert judgements expressed through linguistic variables, and residual subjectivity may persist despite Delphi refinement. Third, the cross-sectional design does not capture the dynamic evolution of governance-quality relationships over time. Finally, the threshold-based filtering emphasizes dominant relationships and may overlook weaker yet meaningful interactions, while implementation contingencies across organizational contexts remain unexamined. Originality/value The study offers one of the first empirical causal mappings linking AI governance and quality assurance. By integrating total quality management principles with AI governance, it reframes governance as a systemic quality capability and provides actionable guidance on sequencing governance investments.